Improvements to LS-DYNA Implicit

LSTC is continually improving the capabilities and performance of Implicit especially in the distributed memory computing environment. This talk will review recent improvements such as (a) removing serial memory restrictions with a new matrix ordering algorithm, (b) reusing the matrix ordering when the matrix structure has not changed from the previous factorization (c) general improvements to performance of the numerical factorization. We will demonstrate these improvements with a series of industrial benchmark problems.